The field of the invention is that of telecommunications networks conforming to the IP Multimedia Subsystem (IMS) architecture defined by the Third Generation Partnership Project (3GPP).
IMS Release 7 defines Policy and Charging Rules Function (PCRF) entities for quality of service and charging control. When a call is set up in the IMS network, a Proxy Call Session Control Function (P-CSCF) entity sends the PCRF a request to communicate information on the application session, the PCRF subsequently using this information to make a quality of service (QOS) and charging decision on activation of the resources for the session.
From this point of view, the P-CSCF entity behaves like an application function (AF), also referred to in this document as an application entity.
On reception of this request the PCRF identifies the transmission links on which the application session impacts and defines a set of policy rules for them.
The above-mentioned standard assumes that during resource activation the termination point in the transport plane, namely the gateway behind which the terminal accesses the IMS network (Gateway GPRS Support Node (GGSN) in a UMTS network, packet data gateway (PDG) entity in an I-WLAN network), selects the PCRF contacted by the application entity AF at the time of setting up the session.
From this point of view the gateway behaves like a Policy and Charging Enforcement Function (PCEF) implementing the QOS and charging policy.
To be more precise, sections 6.2.2.1 and 6.2.2.3 of the document TS 23.203V7.3.0 specify methods of selection of the PCRF by the gateway and by the application entity AF, respectively:                the gateway uses the Packet Data Network (PDN) and information on the identity of the user, for example their IMSI or MSISDN; and        the application entity AF uses the IP address of the user terminal or another identifier of the terminal.        
It is clear to the person skilled in the art that it is necessary for the information used by the gateway and the application entity AF to be identical or well synchronized so that it is certain they contact the same PCRF.
The current standard does not guarantee this. A solution envisaged by certain manufacturers is to use the IP address of the user terminal as the PCRF selection criterion. This solution is unsatisfactory since it does not make it possible to select the same PCRF if the IP address of the user terminal as seen by the gateway is different from that as seen by the application entity AF, especially if the network address is masked by the use of a mobility protocol of the Mobile IP type.
The document WO 2007/090463 describes a method aiming to solve this problem. In that method, the gateway initially selects a PCRF, after which the application function AF selects an available PCRF on the basis of the IP address assigned to the user terminal. If the PCRF selected is already serving the terminal (identified by its IP address), then that PCRF is used for the session. If not, the PCRF selected in error sends back an error message and the application function AF selects another PCRF, again on the basis of the IP address of the terminal, and this process continues until it is the PCRF serving the terminal that is selected.
That method does not guarantee that the application function AF will eventually select the PCRF selected by the gateway if the IP addresses of the terminal as seen by the gateway and the application entity AF are different. It is clear to the person skilled in the art that, PCRF selection being based on the IP address, this method cannot cover the situation in which the operator seeks to retain the same PCRF in mobility scenarios in which the user changes IP address during the same session.
Last but by no means least, that method has the drawback of impacting on the overall performance of the system, especially in operator networks deploying many PCRF entities because, as already stated, the application function AF systematically attempts to select a potential PCRF until the PCRF selected by the gateway is selected.